Large-agency capacity economics without wishful math
Most supervision ROI claims fall apart in the budget office because they count things nobody can promise. Here is the math that holds, run for a 40-officer department, a 150-officer agency and a 600-officer state system.
Budget offices have seen the vendor slide. Recidivism down, jail beds empty, millions saved, all projected from a pilot of twelve people. They have learned to discount it, and they are right to.
So this is the version we would want to defend in front of a finance director. It counts one thing: officer time, which comes out of your own officer count and what a position costs you. Everything else is listed separately as what it is, which is upside you should not build a budget on.
Start from the one number you can defend before day one
Across the agencies we work with, officers get 8 to 10 hours a week back. That is where it comes from:
- A county corrections department ran a time study on its own officers before it ever talked to us. 72.7% of their time went to administrative work. Only 18.7% went to directly supervising people.
- A state department of corrections measured 10.8 hours saved per officer per week in a 1,000-person pilot.
- A statewide juvenile justice agency counted 11,927 officer hours returned in six months, from the platform’s own activity counts.
Before the platform, most of an officer’s week was not supervision
The hours come from the work the platform takes off the calendar: a visit note that took 30 to 45 minutes and now takes about five, monthly reports that stop arriving on paper, the phone tag that a chat message replaces. None of it is a projection about behavior. It is time, counted.
Three agencies, one calculation
The arithmetic is the same at every size. Hours back per officer per week, times 50 working weeks, times the officers on the platform. A position is 2,080 hours a year. The value of a position is whatever a fully loaded officer costs you; we use $76,700, the national median salary from Bureau of Labor Statistics data plus benefits, and you should use your own.
| Same math, three sizes | A county department | A large agency | A state system |
|---|---|---|---|
| Officers on RePath | 40 | 150 | 600 |
| Officer hours back a year | 16,000 to 20,000 | 60,000 to 75,000 | 240,000 to 300,000 |
| Positions of capacity | about 8 to 10 | about 29 to 36 | about 115 to 144 |
| Capacity at $76,700 a position | $590,000 to $737,500 | $2.2 million to $2.8 million | $8.9 million to $11.1 million |
What the number means, and what it does not
This is capacity, not headcount. Nobody cuts a position because the platform came in. The hours go back into the week and the director decides where they go: a caseload that finally gets a home visit, a treatment court that gets its officer back from the scanner, an officer who leaves on time.
It also does not mean the agency wrote a check for $2.8 million less. It means the agency got the work of 29 to 36 officers it did not have to hire. Whether that becomes money depends on what the agency does with it, which is the next section.
The money is usually already in the budget
Most supervision agencies carry vacancies they cannot fill. The salary line for those positions is appropriated and sitting there, and the work is not getting done.
That is where one state corrections department found the money. Going into this year it had about 30 officer positions it could not fill. Its 15-week pilot projected about 13 positions of capacity coming back statewide. The department told us it removed vacant positions and used that salary line to fund a three-year statewide agreement. No new appropriation. The full account is in the case study.
One state’s returned capacity covered 13 of the 30 positions it could not fill
13 of 30vacant positions covered by returned officer capacity, and the salary line that funded the statewide agreement
For the three agencies above, the question is the same: how many of the positions in that capacity row are vacancies you are carrying anyway? That is the number to put in the budget request, because the finance director already knows it is real.
The lines we leave off, and why
There is real money in each of these. We keep them off the budget line because an agency cannot count them before it starts, and a budget request that leans on them is the slide the finance director has learned to discount.
- Missed court dates. A missed court date costs the system about $2,850 and preventing one costs about $10, from a randomized trial of automated reminders published in Science Advances in 2025. The same trial found reminders cut warrants for missed dates by about 20%. The catch: you need your own failure-to-appear rate to size it, and most agencies do not have one they trust. If you do, the calculator will run it. If you do not, leave it off.
- Hardware you stop buying. Ankle monitors and the check-in apps the platform replaces. This one is countable, because it is on an invoice today. Put it in if you have it.
- Revocations and recidivism. One recidivism event costs about $41,450, by the Council of State Governments’ estimate. Across one state, revocations among people on RePath fell from about a third of closed cases in 2022 to 17% last year. That is a trend on the platform, not a controlled study, and we would not put it in front of a finance director as a dollar figure. We would put it in front of a commissioner as the reason the capacity matters.
Run it on your own numbers
Every assumption above is on the calculator, and it needs no email to see the result. Enter your officer count, your loaded cost and the positions you cannot fill, and it gives you the hours, the positions and the dollars the way a budget office needs them. Run the ROI numbers.
If the number on the vacancy line is bigger than you expected, that is the conversation worth having.
Read next
Want to see what your agency could measure?
A 25-minute walkthrough on an agency like yours, with the math on your numbers.
Book a Demo Run the ROI numbers